The UCSC Genome Browser

Transcription

1 The UCSC Genome Browser Donna Karolchik, 1 Angie S. Hinrichs, 1 and W. James Kent 1 UNIT Center for Biomolecular Science and Engineering, University of California Santa Cruz, Santa Cruz, California ABSTRACT The University of California Santa Cruz (UCSC) Genome Browser is a popular Webbased tool for quickly displaying a requested portion of a genome at any scale, accompanied by a series of aligned annotation tracks. The annotations generated by the UCSC Genome Bioinformatics Group and external collaborators include gene predictions, mrna and expressed sequence tag alignments, simple nucleotide polymorphisms, expression and regulatory data, phenotype and variation data, and pairwise and multiplespecies comparative genomics data. All information relevant to a region is presented in one window, facilitating biological analysis and interpretation. The database tables underlying the Genome Browser tracks can be viewed, downloaded, and manipulated using another Web-based application, the UCSC Table Browser. Users can upload personal datasets in a wide variety of formats as custom annotation tracks in both browsers for research or educational purposes. This unit describes how to use the Genome Browser and Table Browser for genome analysis, download the underlying database tables, and create and display custom annotation tracks. Curr. Protoc. Bioinform. 40: C 2012 by John Wiley & Sons, Inc. Keywords: Genome Browser Table Browser human genome genome analysis comparative genomics human variation next-generation sequencing human genetics analysis biological databases BAM bioinformatics bioinformatics fundamentals INTRODUCTION The rapid pace of public sequencing and analysis efforts on vertebrate genomes, combined with the advent of next-generation sequencing, has escalated the demand for tools that offer quick and easy access to the data and annotations at many levels and facilitate comparative data analysis. The University of California Santa Cruz (UCSC) Genome Bioinformatics Web site at provides links to a variety of genome analysis tools, most notably the UCSC Genome Browser (Kent et al., 2002; Dreszer et al., 2012), a graphical tool for viewing a specified region of a genome and a collection of aligned annotation tracks. Another tool on the Web site the UCSC Table Browser facilitates convenient access to the MySql database tables (Karolchik et al., 2004) underlying the Genome Browser annotations. Both browsers support a custom annotation tracks feature that enables users to upload their own data, including nextgeneration sequencing data, for display and comparison. The Basic Protocol describes how to display and navigate among the annotation tracks in a selected region of the Genome Browser, configure the browser tracks image to focus on annotations of interest and optimize comparative analysis, link to external information, and download sequence or annotation data. Support Protocol 1 explains how to create and display a custom annotation track based on the user s own data and set up a Genome Browser session to preserve a group of tracks and settings for later use. Support Protocol 2 provides a basic overview of the UCSC Table Browser, describing the most commonly used functions, how to set up a simple query, and some of the advanced features. The Current Protocols in Bioinformatics , December 2012 Published online December 2012 in Wiley Online Library (wileyonlinelibrary.com). DOI: / bi0104s40 Copyright C 2012 John Wiley & Sons, Inc. Using Biological Databases 1.4.1

2 Genome Browser annotations and software continually evolve as new data and techniques become available; therefore, it is recommended that the user consult the UCSC Genome Browser Web site ( and the current version of the User s Guide ( for the latest information on new releases and features. BASIC PROTOCOL The UCSC Genome Browser USING THE UCSC GENOME BROWSER The Genome Browser software and data may be accessed on the Internet from the UCSC Genome Bioinformatics Web site at Necessary Resources Unix, Windows, or Macintosh workstation with an Internet connection and a minimum display resolution of dpi An up-to-date Internet browser that supports JavaScript, such as Firefox 16.0 or higher ( Internet Explorer 7.0 or higher ( Chrome 22.0 or higher ( or Safari 5.1 or higher ( browser must have cookies enabled Navigate to a specific genomic position in the Genome Browser window 1. Open the UCSC Genome Bioinformatics home page at in a Web browser. The UCSC Genome Bioinformatics home page provides links to the Genome Browser application and a variety of other useful tools: BLAT (Kent et al., 2002; UNIT 10.8), for quickly mapping sequences to a genome assembly; the Table Browser (Karolchik et al., 2004; Dreszer et al., 2012), for viewing and manipulating the data underlying the Genome Browser; the Gene Sorter (Kent et al., 2005), for exploring relationships (expression, homology, etc.) among groups of genes; VisiGene, for browsing through a large collection of in situ mouse and frog images to examine expression patterns; an in silico PCR tool for rapidly searching a sequence database with a pair of PCR primers; and Genome Graphs, a tool for viewing quantities plotted along chromosomes. General information about the Genome Browser tool suite can be found in the User s Guide accessed via the Help link and the FAQ. From the home page, the user can also download genomic sequence and annotation data, browse a collection of contributed custom tracks and older archived data, review a log of released data, and access helpful utilities, training materials, credits for contributors and collaborators, mirror information, and related publications. The home page provides direct links to portals for the Encyclopedia of DNA Elements (ENCODE) Project Consortium (The ENCODE Project Consortium et al., 2004, 2007, 2012; Rosenbloom et al., 2012) and the Neanderthal Genome Analysis Consortium data (Green et al., 2010). 2. Click the Genome Browser link in the left-hand sidebar menu to open the Genome Browser Gateway page. On the Gateway page (Fig ), the user can set the parameters that determine which region of a genome the Genome Browser will initially display. The bottom portion of the page provides information about the currently selected genome assembly and a list of sample position queries that can be used to open the Browser. Alternatively, the Genome Browser can be accessed by clicking on the BLAT link on the home page and then searching a DNA or protein sequence for regions of homology (step 17). 3. Select the clade (or group), genome, and assembly of interest, and then type one or more search terms, a gene name, or a set of genomic coordinates into the Search Term text box to specify the genome region to display. Click the Submit button. Current Protocols in Bioinformatics

3 Figure The Genome Browser Gateway page, set up to span the region of chromosome 4 (chr4: ) in the February 2009 hg19 human assembly (GRCh37) that corresponds to the location of the PHOX2B gene. The display range can be set to the position of a specific gene by typing the name into the Search Term text box. The user can generate a list of tracks containing certain attributes by clicking the Track Search button. Custom annotation tracks (see Support Protocol 1) can be uploaded by clicking the Add Custom Tracks button. The initial Genome Browser display may be configured by clicking the Configure Tracks and Display button. The lower portion of this page (not shown) displays a description of the selected assembly, relevant links, and examples of queries that may be entered in the Search Term box. The search supports direct positional queries such as chromosome bands or chromosome coordinate ranges, as well as queries related to genomic features such as gene symbols, mrna or EST accession numbers, identifiers for single nucleotide polymorphisms (SNPs), author names, or other descriptive terms likely to occur in GenBank (Benson et al., 2011). The Gateway page shows examples of valid position requests applicable to the selected genome assembly. If the position query is resolved to a single location, the Genome Browser will display a page containing an annotation track image specific to the position query, accompanied by navigation controls and display controls (Fig ). Frequently, the position search returns a list of several matches in response to a query rather than immediately displaying the Genome Browser page. When this occurs, click on the item of interest and the Genome Browser will open to that location. Invalid position queries (e.g., withdrawn gene names, abandoned synonyms, misspelled identifiers, and data added after the last Genome Browser database update) will result in a warning message and the previous or default position will be retained. Personal data sets, in the form of custom annotation tracks, can be uploaded into the Genome Browser by clicking the Add Custom Tracks button on the Gateway page. For more information on creating and uploading custom tracks, see Support Protocol 1. To search for one or more specific terms in the entire set of track names, descriptions, and track groups for the current assembly, as well as ENCODE metadata (selected human and mouse assemblies only), click the Track Search button. To access an older genome assembly that is no longer available from the assembly menu, look in the Genome Browser archives at accessible from the Archives link on the home page. Several aspects of the Genome Browser display can be customized by clicking the Configure Tracks and Display button (step 9). Browse and configure the annotation tracks display 4. Explore the Genome Browser annotation tracks page (Fig ). This image displays a set of annotation tracks aligned beneath a Base Position track (the ruler ) indicating genomic coordinate positions in a 5 (left) to 3 (right) orientation. Tracks are organized into groups reflecting the nature of their data. The first time the Genome Browser is opened, the application s default values are used to configure this display. Any preferences and configurations set during the session will be retained for use in subsequent sessions on the same Web browser. To reset the display to the set of default tracks for the selected assembly, click the Default Tracks button. Using Biological Databases Current Protocols in Bioinformatics

4 Figure The Genome Browser annotation track page zoomed out to display the PHOX2B gene and its 5 and 3 flanking regions on human chromosome 4 (chr4: ) in the Feb assembly (GRCh37/hg19). The navigation and configuration buttons are visible above and below the image. The red rectangle in the ideogram directly above the annotation tracks image indicates the location of the currently displayed region of the chromosome. The Common SNPs (132) track visibility has been changed from dense to pack to show individual SNPs, some of which are colored according to gene region (e.g., UTR or coding-nonsynonymous), and the Flagged SNPs (132), showing SNPs flagged as clinically associated in dbsnp, has been added by changing the display visibility to pack. The Phenotype and Disease Associations group has been opened, and three tracks from the group have been added to the display by changing their visibilities from hide to pack: GAD View, OMIM Genes, and OMIM AV SNPs. PHOX2B is a developmental gene that has also been associated with cancer; move the mouse over the PHOX2B item in the GAD View track to see a list of diseases associated with the gene. In the Vertebrate Multiz Alignment & Conservation track, note the areas of high conservation peaking in the upstream region (to the right because PHOX2B is on the antisense strand), UTRs, and most exons, as well as part of the first intron. For the color version of this figure, go to The UCSC Genome Browser The complete set of available annotation tracks for the assembly is shown in the track groups section below the image, categorized by data type. Many of the tracks on the later human genome assemblies were contributed by the ENCODE Project; these are denoted by a double helix icon in the track label. Tracks generated by simply converting ( lifting ) the coordinates of the data from a previous assembly are marked by a black circle on which the UCSC version number of the originating assembly has been superimposed. See the Commentary section for more discussion of the annotation tracks available on the human genome. The annotation tracks image is accompanied by control buttons to configure the display and navigate through the sequence. For selected assemblies, a chromosome band ideogram directly above the image graphically indicates the location of the currently displayed region on the chromosome. Personal annotation data can be uploaded to the current assembly by clicking the Add Custom Tracks button below the image (see Support Protocol 1 for more information). Figure shows the annotation track image opened to the position of the gene PHOX2B on chromosome 4. To reach this position, enter PHOX2B in the gene text box, click Submit, and then click the zoom out 1.5 button. (This location may also be reached by typing PHOX2B in the search box, clicking Submit, and then selecting the first matching item the UCSC Genes PHOX2B.) Note that the Genome Browser automatically changes Current Protocols in Bioinformatics

5 the text in the Position box to show the chromosomal position of the resulting display, in this case chr4: In most annotation tracks, the aligned regions are represented by vertical bars or blocks. In the Spliced ESTs track shown in this example, the degree of darkness of the block shading corresponds to the number of features aligning to the region. In the mrna and gene prediction tracks, the thicker regions (usually coding exons) are connected by thin horizontal lines representing gaps (usually spliced-out introns). Thinner blocks on the leading and trailing ends of the aligning regions in gene tracks represent the 5 and 3 untranslated regions (UTRs). In full or pack display mode, arrowheads on the connecting lines indicate the direction of transcription. Note the comparative genomics annotations displayed in Figure The Conservation track shows a measure of evolutionary conservation among multiple species, which tends to indicate functional regions of the genome. The lower section of the track shows pairwise alignments of each species to the reference sequence; the top section displays the evolutionary conservation scores assigned by the phylop (Pollard et al., 2010) and phastcons (Siepel et al., 2005) methods (hidden by default) in the PHAST package (Siepel et al., 2005). In human assemblies, the display may be configured to show primates, placental mammals, or other vertebrates. When displayed in full or pack mode, the conservation track is a good example of wiggle (histogram) format in which the height reflects the magnitude of the score. At the level of detail shown in Figure 1.4.2, the scores highlight exons, untranslated regions (UTRs), and other regions that show signs of conservation across species. To generate a high-quality image of this annotation tracks image in PostScript or PDF format, click the PDF/PS link in the View menu on the top menu bar. 5. Change the display mode of an annotation track by right-clicking on it in the track image and selecting a display mode from the menu that displays, or by locating the track s name in the track group section below the image, selecting a display mode from the track s pull-down menu, and then clicking the Refresh button. Depending on individual display modes, annotation tracks may be hidden from view (hide mode), displayed with all features collapsed into a single line (dense mode), or fully expanded with each feature on a separate line (full mode). Many tracks feature two additional display modes: pack mode, in which each feature is displayed and labeled, but not necessarily on a separate line, and squish mode, which is similar to pack mode, but displays unlabeled features at half-height. To quickly toggle between dense and full (or pack) modes for a displayed track, click on the track s label in the annotation track image. To hide all the tracks in the display, click the Hide All button beneath the annotation tracks image. By adjusting the display modes of the tracks in the annotation track graphic, the user can restrict the display to data of interest, reduce clutter, and improve speed. Dense display mode is useful to get an overview of an annotation or to reduce the space used by a track when the individual feature details of an annotation track are not required. Squished and packed displays show individual feature details of densely populated tracks while conserving space. Use full mode sparingly; in some tracks, the number of features that may potentially align at a selected position can be quite large. When the feature count is excessive in full display mode, the Browser displays the track in pack mode if possible; if the track does not support pack mode, it displays the first 250 items individually, then groups the remaining items into a single line in dense mode at the bottom of the track. 6. To move to a different genomic position, type a new set of search terms into the position/search box or a different gene name into the gene text box, then click the Jump button. Figure illustrates a zoomed-in view of the genomic region displayed following a search on the SNP identifier rs on the Feb (GRCh37/hg19) human assembly. This is a common coding nonsynonymous SNP in the human gene CYP4F2 that has been associated with warfarin dosage responsiveness by multiple studies in the NHGRI catalog of published genome-wide association studies (GWAS) (Hindorff et al., 2009). Additional tracks in the Variation and Repeats group and the Phenotype Using Biological Databases Current Protocols in Bioinformatics

6 Figure The Genome Browser view displaying the bases surrounding the SNP rs on chr19 (chr ) in the Feb human assembly (GRCh37/hg19). To view this region, enter rs in the search box, choose one of the three results in the GWAS Catalog track, and then click the Base button in the Zoom navigation section on the annotation track page. This zooms in the display to a level where single-nucleotide bases can be studied; note the bases (A, C, G, T) drawn in the Base Position track, Conservation track, and HapMap SNPs. The orange numbers above the multi-species alignment in the Conservation track give the number of bases present in the other species, but not in the human reference, where an orange tick mark appears below. In the HapMap SNPs track, the major allele in each population is displayed instead of the usual colored box. For the color version of this figure, go to The UCSC Genome Browser and Disease Association group have been opened (step 5) to search for supplemental information. The SNP annotation tracks in the Genome Browser show mappings of single-nucleotide polymorphisms and small insertions and deletions (indels) from the dbsnp database (Sayers et al., 2012). The dbsnp build 132 data have been separated into four distinct Genome Browser tracks to facilitate study. The Human Genome Diversity Project (HGDP; and HapMap (The International HapMap Consortium, 2003, 2005; The International HapMap 3 Consortium, 2010) tracks show SNPs genotyped in several populations worldwide. The Genetic Association Database (GAD; Becker et al., 2004) and GWAS catalog tracks show data from human genetic and genome-wide association studies. The Online Mendelian Inheritance in Man (OMIM; Amberger et al., 2009, 2011; Baxevanis, 2012) tracks show variants in the OMIM database that have been associated with dbsnp identifiers, the genomic positions of gene entries in the OMIM database, and regions known to be associated with a phenotype, but lacking a known causative gene. (See the Commentary section for a broader discussion of the annotation tracks in these two groups as well as cautions for using the data.) Current Protocols in Bioinformatics

7 Figure The Genome Browser annotation track page displaying chromosome bands 22q13.32 and 22q13.33 on chromosome 22 (chr22: ) in the Feb human assembly (GRCh37/hg19). Several tracks useful for the display of large regions have been made visible: from the Mapping and Sequencing Tracks group, Chromosome Bands and Gap; from the Phenotype and Disease Associations, GAD View, OMIM Genes and RGD Human QTLs; and from the Variation group, Flagged SNPs (132), Mult. SNPs (132), HGDP Allele Freq, HapMap SNPs, and DGV Structural Variation. Squish display mode (see Basic Protocol, step 5) has been set for UCSC Genes and DGV structural variation to show the density of items in those tracks along the genome. Several tracks have been hidden because they have so many items in this large region that they would display as solid bars in dense mode, or take up large amounts of vertical space if displayed in pack or squish mode. Figure shows a larger region obtained by entering the query 22q13.32; 22q13.33 on the Feb (GRCh37/hg19) human genome assembly. Several tracks that display best in large regions due to the sparseness of their annotations have been added to the display, and several tracks whose many items would saturate the display have been hidden. At this broad scale, the completeness of the assembly is indicated by the sparse gaps. It is easy to see regions of relative gene density or scarcity. Coarse measures such as population genetic statistics have more of a perceivable signal, while fine-scale measures such as the per-base Conservation scores have almost no signal due to averaging over large numbers of bases. 7. Use the mouse drag-and-zoom and drag-scroll features or the Zoom and Move buttons to increase or decrease the breadth of the displayed coordinate range, or to shift one or both ends of the coordinate range to the left or right. To quickly zoom in to an exact coordinate range, click on the desired leftmost coordinate in the Base Position track and drag the mouse to the right to highlight the region of interest. Scroll the tracks image horizontally by clicking on the image, dragging the cursor to the left or right, then releasing the mouse button. The navigation buttons are useful for Using Biological Databases Current Protocols in Bioinformatics

8 generally focusing the display on a position. Zoom buttons increase or decrease the displayed coordinate range by 1.5-, 3-, or 10-fold. To zoom in by 3-fold on a particular coordinate, click the Base Position track at that location. To rapidly zoom in to the base composition of the sequence underlying the current annotation track image, click the zoom-in Base button. Move buttons shift the displayed coordinates in the indicated direction by 10%, 50%, or 95% of the displayed size. To scroll the coordinate position of one side of the track display while holding the position of the opposite end static, click the corresponding Move Start or Move End arrow button. For example, to preserve the left-hand display coordinate but increase the right-hand coordinate, click the Move End forward arrow. To increase or decrease the scroll interval, edit the number in the Move Start or Move End text box. The UCSC Genome Browser Use the drag-reorder feature to change the vertical ordering of the tracks. To move a track up or down within the tracks image, click-and-hold the mouse button on the side label or gray button to the left of the track, then drag the highlighted track up or down within the image. Release the mouse button when the track is in the desired position. To restore the default ordering of the tracks, click the Default Order button below the tracks image. 9. Click the Configure button below the annotation tracks image to access a Web page for changing display characteristics (such as the image width and text size), hiding, showing, or reordering track groups, and displaying the chromosome ideogram, the track groups section, and image labels. Click the Submit button on the configuration page to apply the changes and return to the annotation tracks image. The Next/Previous Item Navigation and Next/Previous Exon Navigation features provide a quick way to move forward or backward among features or exons in a track. When these features are enabled, gray or white double-headed arrows will display on the 5 and 3 sides of tracks supporting the feature. Clicking on an arrow shifts the image view window toward that end of the chromosome so that the next item or exon in the track is displayed. The default display width of the annotation track graphic is optimized to best fit the size of the Web browser window the first time the tracks Web page is viewed. If the Web browser window is subsequently resized, the annotation track image size can be automatically readjusted to the new width by clicking the resize button under the track image. The default width can still be manually overridden on the Track Configuration page. Exercise caution when using the Show All option in the track configuration section: if the group or assembly has a large amount of annotation data, the Web browser session may freeze or terminate before the datasets are loaded. 10. Click the gray button to the left of a displayed track to view additional information about the annotation and (in many cases) to filter or configure the features displayed in the track. The description page can also be displayed by clicking the track s name in the track groups section. Click the button adjacent to the UCSC Genes track to view a typical description page. This page contains a configuration/filter section (when applicable) followed by a description of the annotation track, information about interpreting and configuring the track display, a discussion of the methods used to collect and compute the data, credits for authors and contributors, associated references, and in this case, restrictions on the use of the data. Additional credits can be found by clicking the Credits link on the home page. Most of the tracks in the Genome Browser have filter or configuration options that modify the graphical characteristics or restrict the display to features that match filtering criteria. Filters are useful for focusing attention on relevant features when a track contains large amounts of data. Some of the more complex graphical annotations, such as the continuous value graph ( wiggle ) display featured in the Conservation track, offer an extensive set of configuration options. In most cases, detailed configuration information can be found in the Display Conventions and Configuration section on the description page. Current Protocols in Bioinformatics

9 Filter and configuration settings are persistent from session to session on the same Web browser. To revert to the original default settings for a track, manually restore the settings on the description page; to undo all changes that have been made to default settings for any track or tool, click the Click Here to Reset link on the Gateway page. 11. Click on a feature name in a track shown in pack or full display mode to view detailed information about the feature and access links to additional information. The types of information available vary by track. The RefSeq Genes track (Pruitt et al., 2012) provides an example of a typical feature information page. Enter HOXA1 into the gene text box and click the jump button. In the track image, click on the HOXA1 gene label in the RefSeq Genes track to display the associated information page (adjustment of the track display mode may first be necessary step 5). By contrast, the UCSC Genes track has a more extensive collection of information about the gene, including the associated UniProt (The UniProt Consortium, 2011) and RefSeq descriptions, microarray expression data, links to associated information about this gene in several UCSC tools (such as the Gene Sorter and Table Browser), as well as links to related records in external databases, including Online Mendelian Inheritance in Man (OMIM; Amberger et al., 2009, 2011; Baxevanis, 2012), Entrez Gene (Sayers et al., 2012), Ensembl Genes (Flicek et al., 2012), GeneLynx (Lenhard et al., 2001), GeneCards (Safran et al., 2010), AceView, PubMed (Sayers et al., 2012; UNIT 6.10), the HUGO Gene Nomenclature Committee Database (HGNC; Seal et al., 2011), the Cancer Genome Anatomy Project (CGAP; Strausberg et al., 2001), PDB (Rose et al., 2011; UNIT 1.9), ModBase (Pieper et al., 2011), InterPro (Hunter et al., 2009), Pfam (Punta et al., 2012; UNIT 2.5), the Stanford SOURCE (Diehn et al., 2003), Mouse Genome Informatics at Jackson Laboratory (Blake et al., 2011), the Allen Brain Atlas (Lein et al., 2007), and more. The page also includes links that will display the corresponding protein, mrna, and genomic sequences for HOXA1. These sequences are a useful source of input into the BLAT tool, which is discussed in step 17. The Genome Browser also provides direct links to the Ensembl Browser (Flicek et al., 2012; UNIT 1.15) and NCBI s Map Viewer (Sayers et al., 2012; UNIT 1.5), when available. To view the complementary annotation in one of these browsers, return to the annotation tracks page and click the Ensembl or NCBI link in the top menu bar. Examine underlying data and download sequence and annotation data tables 12. To view the DNA sequence underlying an item in the image, right click on the feature and select the Get DNA for... menu option. To view the DNA sequence of the region spanned by the image, click the DNA link in the View menu on the annotation tracks page menu bar. This DNA utility allows the user to change the formatting and coloring of the text that represents the sequence to highlight features of interest. The initial display window provides options for marking or masking repeats, changing the case of the letters that represent the DNA, showing the reverse complement of the sequence, and displaying additional sequence upstream or downstream of the selected sequence. Click the Extended Case/Color Options button to display additional font and color configuration options. The Extended DNA Case/Color Options page is useful for highlighting features within a genomic sequence, pointing out overlaps between two types of features, or masking out unwanted features. In Figure 1.4.5, the configuration has been set to display exons from the UCSC Genes track in uppercase letters. The Spliced EST track is configured to reflect the level of coverage by setting its color to RGB value (0, 64, 0). Common SNPs will display in a bold font, and Flagged SNPs will be underlined. When the Submit button is clicked, the Extended DNA Output window shown in Figure is displayed. Note that only tracks currently visible in the Genome Browser tracks image are available for configuration on the Extended DNA Case/Color Options page. Be careful when requesting complex formatting for a large chromosomal region: when all the HTML tags have been added to the output page, the file size may exceed the limits that the Web browser, clipboard, and other software can display. Using Biological Databases Current Protocols in Bioinformatics

10 Figure An extended DNA Case/Color Options request to display the DNA for the chr4: region of the Feb (GRCh37/hg19) human assembly. This configuration sets up a display that will show UCSC Genes in uppercase, all other regions in lowercase, and Spliced ESTs in varying intensities of green, depending on the level of coverage. Common SNPs are shown in bold, and Flagged SNPs are displayed in bold and underlined. Figure Output from the DNA display configurations set up in Figure Exons are shown in uppercase. Nucleotides covered by a single EST appear darker green on the screen, while regions with more EST alignments appear progressively brighter, saturating at four ESTs. Common and Flagged SNPs are called out. For the color version of this figure, go to The UCSC Genome Browser Click on the Table Browser link in the Tools menu on the annotation tracks page menu bar to access the database tables underlying the Genome Browser annotation tracks. The Table Browser tool provides a graphical interface for viewing and manipulating Genome Browser data. Support Protocol 2 gives a brief introduction to using the Table Browser. Additional information can be found in the Table Browser User s Guide accessible from the Help link in the Table Browser top menu bar. Current Protocols in Bioinformatics

11 14. Click the Home link on the top menu bar to return to the UCSC Genome Bioinformatics home page, and then click the Downloads link on the side bar to display a listing of sequence files and database tables available for downloading. The Downloads page contains links to all the Genome Browser assemblies, annotations, and source code available on the Genome Browser downloads server. To access older assembly versions, it may be necessary to look in the archives ( Data is also downloadable at the Genome Browser FTP site (ftp://hgdownload.cse.ucsc.edu/goldenpath/). FTP or rsync is recommended for large data downloads. All data in the Genome Browser are freely available, except where noted in the README.txt file specific to a particular downloads directory. The Genome Browser and BLAT source are freely available for academic, noncommercial, and personal use; commercial licensing information can be found via the Licenses link on the home page. Convert coordinates in the displayed range to a different assembly using the Convert, LiftOver, or BLAT tools 15. Return to the annotation tracks page. Click the Convert link in the View menu on the menu bar to convert the coordinates in the displayed range to those of a different assembly. The coordinate conversion tool is useful for locating the position of a feature of interest in a different genome assembly. Coordinates of features frequently change from one assembly to the next as gaps are closed, strand orientations are corrected, and duplications are reduced. For example, to map the location of a sequence in the hg18 (Mar. 2006) human assembly to the hg19 (Feb. 2009) human assembly, open the hg18 Genome Browser to the desired position, click the Convert link in the View menu, select the hg19 option in the New Assembly pull-down menu, then click the Submit button. If successful, the Convert tool displays one or more coordinate ranges in the hg19 assembly to which the hg18 sequence maps. 16. To convert multiple sets of sequence coordinates between assemblies or to exert control over the parameters used in the conversion, use the LiftOver batch coordinate conversion tool. The LiftOver tool can be accessed from the Utilities link on the Genome Bioinformatics home page or the Tools menu on the annotation tracks page. Enter the list of coordinate ranges in the large text box, one per line, or upload the list from a file. Detailed information about parameter settings can be found at the bottom of the page, as well as information about a Linux command-line version of the tool. 17. Alternatively, use BLAT to map a sequence to a different assembly: a. Obtain the DNA sequence for the region or feature using the methods outlined in step 12. Note that BLAT limits input to 25,000 bases. b. Using the Web browser s copy function, copy the entire sequence onto the clipboard. Return to the annotation tracks page and click the Blat link in the Tools menu on the top menu bar. c. On the BLAT Web page, paste the sequence into the large text box (Fig ). Select the genome and assembly to which to map the sequence, and then click the Submit button. If successful, BLAT will display a list of search results sorted by score (Fig ). d. To view the details of the matching alignments, click the Details link; to display the sequence in the Genome Browser, click the Browser link. This procedure demonstrates one use of the BLAT search tool. This tool, which can be accessed from the BLAT link in the Tools menu on the top menu bar of most Genome Browser pages, is a very fast sequence alignment tool similar to BLAST (UNIT 3.3), but optimized for inputs with high similarity, e.g., sequences from the same species. For more information on BLAT, refer to the Genome Browser User s Guide. Using Biological Databases Current Protocols in Bioinformatics

12 Figure A BLAT search set up to align the FASTA sequence in the text box against the Feb (GRCh37/hg19) human genome assembly. This sequence was obtained by copying and pasting the output from the Get DNA search illustrated in Figures and Figure The results returned by the BLAT search shown in Figure Clicking on the Browser link for a given line will display the data in the Genome Browser; the Details link will display a page showing a base-by-base of the alignment to the genome. SUPPORT PROTOCOL 1 The UCSC Genome Browser CREATING A CUSTOM ANNOTATION TRACK Custom annotation tracks enable users to upload personal data for temporary use in the Genome Browser and Table Browser. Custom tracks are viewable only on the machine from which they are uploaded, and by default the data may be accessed only by the users on that machine. Optionally, users can make custom annotations viewable by others through the use of Genome Browser sessions or custom track URLs. Tracks are kept for 48 hr after the last time accessed unless they are saved in a Genome Browser session; no permanent archives are created. The Genome Browser custom track feature accommodates user-generated data in a wide variety of formats. Smaller datasets may be structured using one of the formats developed during the early years of the Human Genome Project, such as general feature format (GFF), gene transfer format (GTF), pattern space layout (PSL), and browser extensible data (BED), or a format developed for special Browser display purposes, such as wiggle (WIG) and bedgraph formats for continuous-valued data, multiple alignment format (MAF), microarray (BED15) format, and personal genome SNP format for displaying variant base calls from personal genomes relative to the reference genome. Current Protocols in Bioinformatics

13 The larger datasets that have become more common with next-generation sequencing and whole-genome analysis usually require a compressed, indexed format to avoid potential performance issues and Internet timeout problems associated with large data file uploads. Formats supported by the Genome Browser include bigbed and bigwig (Kent et al., 2010), which are the indexed binary format versions of the BED and WIG formats, and Binary Alignment/Map format (BAM), the compressed binary version of the Sequence Alignment/Map (SAM) (Li et al., 2009) format used to represent the alignment of nextgeneration nucleotide sequencing reads to a reference genome. Support was added recently for Variant Call Format (VCF; Danecek et al., 2011), a variation data interchange format initially developed for the 1000 Genomes Project (1000 Genomes Project Consortium, 2010) to display SNPs, indels, copy number variations (CNVs), and structural rearrangements. The Genome Browser can display VCF files that have been compressed and indexed using tabix (Li, 2011). Genome Variation Format (GVF; Reese et al., 2010) is the format chosen by the Database of Genomic Structural Variation (dbvar; Sayers et al., 2012) to encode hierarchical structural variants. When custom tracks using the indexed binary formats are loaded into the Browser, both the track file and its associated index file remain on the user s Web-accessible server (http, https, or ftp), and only the portions of the files needed to display a particular genomic region are transferred to UCSC where they are temporarily cached. Typically, custom annotation tracks are displayed under the corresponding genomic positions on the Base Position track. Each custom track has its own track control and persists even when not displayed in the Genome Browser window (e.g., if the position changes to a range that no longer includes the track). Once displayed, a custom track can be moved up or down in the tracks display just like standard Genome Browser tracks. Custom tracks can be saved for later use through the Genome Browser Session tool, which allows a user to preserve a specific set of Browser track combinations and configuration options. Multiple sessions may be saved for future reference, for comparing different data sets, or for sharing with colleagues. Saved sessions persist for 4 months after the last access or until deleted. Since space is limited in the annotation track graphic, many excellent genome-wide tracks must be excluded from the set provided with the Browser. A Web page with links to user-contributed custom tracks can be found by clicking the Custom Tracks link on the home page. In 2011, the Genome Browser expanded its display support of personal data to include track data hubs, which are user-hosted Web-accessible directories of genomic data that can be viewed in the Browser alongside the native annotation tracks (Dreszer et al., 2012). Track data hubs allow researchers to combine and configure large numbers of datasets for presentation as a single entity, while taking advantage of the display performance achieved through the use of indexed binary format. A track hub may be shared privately with specific colleagues or publicly with the general research community. For more information on constructing and using track data hubs, see The information in this section provides an overview of the process for creating and displaying custom annotation tracks in the Genome Browser. For a more detailed discussion of formats, syntax, and utilities, refer to the Genome Browser custom annotation track documentation Web page at html. Using Biological Databases Current Protocols in Bioinformatics

14 Necessary Resources Unix, Windows, or Macintosh workstation with an Internet connection and a minimum display resolution of dpi Text editor (APPENDIX 1C) An up-to-date Internet browser that supports JavaScript, such as Firefox 16.0 or higher ( Internet Explorer 7.0 or higher ( Chrome 22.0 or higher ( or Safari 5.1 or higher ( browser must have cookies enabled Display a small dataset as a custom track 1. Format the dataset to be analyzed as a tab-separated file using one of the plain text formats supported by the UCSC Genome Browser: GFF, GTF, PSL, BED, WIG, bedgraph, MAF, BED15, or personal genome SNP. Each data line in a plain text file provides display and positional information for an item within the displayed annotation track. The Browser ignores empty lines and lines starting with a pound sign (#). Data in PSL, GFF, and GTF files must be tab-delimited rather than space-delimited in order to display correctly. More than one dataset may be included in an annotation file, but all lines within a single annotation track must be in the same format. An easy way to create correctly formatted data for an annotation file is by collecting PSL output from BLAT or downloading data from the Table Browser. Figure shows examples of data in BED, PSL, and GFF format. For detailed information on custom track data formats, refer to the Genome Browser s custom annotation track documentation and file format FAQ ( 2. Add one or more optional browser lines to the beginning of the formatted data file to specify the configuration of the Genome Browser window in which the custom annotation track will be displayed. The UCSC Genome Browser Figure Sample custom annotation tracks containing BED, PSL, and GFF data formats. To load correctly, the track line data in the PSL and GFF examples must be tab-separated. Some of the line breaks shown in the BED and PSL examples are artificial (to make the text fit on the page); browser, track, and data lines may not contain internal line breaks. Current Protocols in Bioinformatics

15 Browser lines define the genome position to which the Browser will initially open, the width of the display, and the configuration of the other annotation tracks that are shown (or hidden) in the initial display. The Genome Browser custom annotation track documentation describes the browser line syntax and options. In the sample BED annotation track shown in Figure 1.4.9, the initial display position is set to chr22: , and all tracks are hidden except the custom annotation track. If the browser position is not explicitly set in the annotation file, the initial display will default to the position setting most recently used by the user, which may not be an appropriate position for viewing the annotation track. 3. Add a track line immediately above the formatted data in the file to define the display attributes for the annotation track. The track line defines the track s name, description, colors, initial display mode, associated URL, and other settings. The Genome Browser custom annotation track documentation contains a complete description of the track line syntax and options. If more than one dataset is included in the annotation file, insert a track line at the beginning of each new set of data. In Figure the left-hand label of the BED annotation track is BED track; the center label is BED track example. The track labels will be displayed in green and the features will be fully displayed. Because the usescore attribute is set to 1, the level of shading of each feature will reflect its score value. 4. Upload the annotation file into the Genome Browser by clicking the Add Custom Tracks button on the Gateway page (Fig ) or the Add Custom Tracks button on the annotation tracks page (Fig ). If the file is located on the same machine on which the Genome Browser is being viewed, enter the file name in the Upload text box in the URLs or Data section. To open an annotation through a URL or to manually enter the track data, type or paste the information into the large text box in this section. Multiple tracks may be uploaded simultaneously by including all the track data or URLs (on separate lines) in the text box or grouping the tracks into one uploaded file. Figure shows the custom track that displays when the BED sample track in Figure is uploaded into the Genome Browser. Optionally, associated track descriptive text may be uploaded or inserted into the Optional Track Documentation section. To make the annotation file viewable on a different machine or at a different site, put a copy of the file on a Web server and create a custom annotation track URL that allows the file to be uploaded over the Internet. The URL must contain two pieces of information specific to the annotation data file: the UCSC genome assembly on which the annotation is based and the URL of the annotation file on the Web site. The Genome Browser FAQ ( lists the UCSC genome assembly codes. The URL can also include the position within the genome to which the Genome Browser should initially open. For example, placing the BED track in Figure in a file named test.bed on the genome.ucsc.edu Web site enables it to be uploaded using the following custom annotation track URL: Figure The annotation track that displays when the BED track example in Figure is uploaded into the Genome Browser. Note that the lower score value in the ItemB data results in lighter shading of this feature Current Protocols in Bioinformatics

16 This URL sets the assembly database to the hg19 (Feb. 2009) assembly of the human genome, initializes the display position to chromosome 22, and loads the annotation track file In this case, the position initialization in the URL is extraneous; it will be overwritten by the position defined in the custom track file. Create a compressed indexed custom track for a very large dataset 5. Convert a very large dataset to one of the compressed formats: bigbed, bigwig, BAM, or VCF. Datasets in BED or WIG format can be converted to bigbed or bigwig format using the Genome Browser bedtobigbed or wigtobigwig utilities. Sequence alignments in SAM format can be converted to BAM format and indexed through the use of SAMtools ( Similarly, variant calls in VCF can be compressed and indexed using tabix ( Consult the Genome Browser help pages on BAM format ( and VCF ( for detailed instructions on using these tools. 6. Move the converted data file and resulting index file to a Web-accessible http, https, or ftp location. 7. Construct a custom track file that defines attributes of the annotation track. At a minimum, the file must contain a single track line that defines the track type attribute and specifies a bigdataurl pointing at the Web location of the compressed data file, for example: track type = bam bigdataurl = Several other optional attribute settings may be defined for bigbed, bigwig, BAM, and VCF format custom tracks. Click on the link for the format of interest on the Genome Browser custom track help page ( hgtrackshelp.html#customtracks) for an in-depth description of the format attributes and examples of their use. One or more optional browser lines may be included before the track line to specify the configuration of the Genome Browser window in which the custom annotation track will be displayed (step 2). Figure shows a custom track file that was constructed to load a small section of a large dataset from the 1000 Genomes Project. This project consortium, whose aim is to discover, genotype, and provide accurate haplotype information on all forms of human DNA polymorphism in multiple human populations, has developed several tools for processing and displaying next-generation sequencing data. The BAM dataset in this example is from the chromosome 21 region of the hg18 human assembly and is available on the Web site specified by the bigdataurl attribute. The UCSC Genome Browser Figure An example of a custom annotation track definition for an indexed BAM file that resides on the NCBI FTP server specified by the bigdataurl attribute. The line breaks are artificial (to make the text fit on the page). No data lines follow the track definition line because the data are retrieved (as needed) from the remote BAM file named in the bigdataurl setting. BigWig, BigBed and tabix-indexed VCF custom tracks have a similar structure. Current Protocols in Bioinformatics

17 Figure The track display of the uploaded BAM format custom track file shown in Figure Load the custom track and display it in the Genome Browser. Click the Add Custom Tracks button on the Gateway page (Fig ) or the Add Custom Tracks button on the annotation tracks page (Fig ) to display the custom track management page. Load the file containing the track line and optional browser lines, then click submit to display the track in the Genome Browser (Fig ). Create a Genome Browser session 9. Create a Genome Browser session that preserves a snapshot of the custom track for future use and for sharing with colleagues. The Session utility enables the saving and loading of customized views of specific genomic regions with selected tracks displayed, including both standard and custom annotation tracks, which can be shared as text files or URLs, or ed to others (see Suggestions for Further Analysis section for more information). To create a session containing a custom track, load the track into the Genome Browser, then click the Session link in the menu bar. A login to the genomewiki system is required before a session may be created. Create a named session under the Save Settings section, specify if the session may be viewed by others, then click Submit to add the session to the My Sessions list. The session containing the custom track may then be loaded into the Browser or sent to colleagues. Refer to the Session User s Guide ( for more information about creating and using Genome Browser sessions. USING THE UCSC TABLE BROWSER The UCSC Table Browser provides a powerful and flexible graphical interface for querying and manipulating the data in the Genome Browser annotation database. The Table Browser can be used to: (1) retrieve the annotation data or DNA sequence underlying Genome Browser tracks for the entire genome, a specific coordinate range, or a set of accessions; (2) view a list of the tables affiliated with a particular Genome Browser track; (3) view the schema of an annotation table; (4) organize table data into formats that can be used in other applications, spreadsheets, or databases; (5) combine data from multiple tables or custom tracks into a single set of output data; (6) filter out certain records in a table based on certain field values; (7) display basic statistics calculated over a selected range of table data; and (8) conduct structured or free-form SQL queries on the annotation data. SUPPORT PROTOCOL 2 Using Biological Databases Current Protocols in Bioinformatics

18 The information in this section provides an overview of the Table Browser, which can be accessed on the Internet from the UCSC Genome Bioinformatics home page at genome.ucsc.edu. For a more detailed discussion of Table Browser options, advanced queries, and several practical examples, refer to the Table Browser User s Guide at For complex queries of Genome Browser data, the Galaxy interactive genome analysis tool ( galaxy.psu.edu/; Blankenberg et al., 2010; Goecks et al., 2010; UNIT 10.5) may be used. Necessary Resources Unix, Windows, or Macintosh workstation with an Internet connection and a minimum display resolution of dpi An up-to-date Internet browser that supports JavaScript, such as Firefox 16.0 or higher ( Internet Explorer 7.0 or higher ( Chrome 22.0 or higher ( or Safari 5.1 or higher ( browser must have cookies enabled Set up a simple Table Browser query 1. On the UCSC Genome Bioinformatics home page (see Basic Protocol, step 1), click the Table Browser link in the left-hand sidebar menu to display the Table Browser Web page. The Table Browser is also accessible from the Tables link in the Tools menu on the top menu bar of most Genome Browser pages. The top section of the Table Browser Web page (Fig ) contains options for setting up a data query, many of which are optional when conducting simple queries. Each of the options is briefly described at the bottom of the Web page. To view the complete Table Browser User s Guide, click the Help link in the top menu bar. Figure The Table Browser tool provides access to the database tables underlying the Genome Browser annotations; in this case, the chromosome 7 data in the knowngene table on the Feb human genome assembly (GRCh37/hg19) Current Protocols in Bioinformatics

19 2. Select the clade (or group), genome, and assembly. The clade (or group), genome, and assembly pull-down menus correspond to those found on the Genome Browser Gateway page. The current Genome Browser settings are used when the Table Browser is started from the menu bar on a Genome Browser page. For this example, set the clade to Mammal, the genome to Human, and the assembly to Feb (GRCh37/hg19). 3. Select the group, track, and table of interest. The options in the group and track menus directly correspond to the annotation groups and tracks available in the Genome Browser for the currently selected genome assembly, including currently loaded custom tracks (see Support Protocol 1). The track list which shows all tracks contained in the selected group automatically updates when a different group is selected. The table menu lists all the tables in the annotation database that are affiliated with the selected track. Many annotation tracks are based on data from multiple tables joined by common fields. By default, the primary table underlying the track s display in the Genome Browser is listed first. Click the Describe Table Schema button to view the SQL schema for the selected table. The schema page also lists other tables in the annotation database that are joined to the selected table by a particular field, as well as a description of the Genome Browser annotation track associated with the table (when applicable). The All Tracks and All Tables options in the group menu provide convenient shortcuts if the name of the desired track or table is already known. For this example, a subset of data in the UCSC Genes track will be examined. Select the Genes and Gene Prediction Tracks group, the UCSC Genes track, and the knowngene (default) table. 4. Specify the query region. Click the Genome region setting to view annotation data for the entire genome. To limit the data output to a specific query region, click the Position region setting and type a query into the adjacent text box. The Table Browser accepts the same types of queries that are valid for the Genome Browser (see Basic Protocol, step 3). Click the Lookup button to convert a nonpositional query (e.g., an accession or keyword) to a coordinate range. On the hg17 and hg18 genome assemblies, which have annotations specific to the ENCODE pilot project, an additional ENCODE Pilot Regions setting is available that restricts output to data located in the 44 ENCODE Pilot regions. There is no need to use this setting for the genome-wide production phase ENCODE data found on hg18 and later assemblies. For many tables, the query region can be further defined by restricting the output to a set of specific identifiers, such as UCSC Gene IDs, mrna accession numbers, or dbsnp IDs. Upload the identifiers as a space- or line-separated list by clicking the Paste List or Upload List button. For this type of query to return successfully, the identifiers in the list must conform to the format specified for identifiers in the selected table. For this example, several UCSC Gene identifiers from chromosome 7 are included in the query. Select the Position region setting, then type chr7 in the text box. Click the Paste List button, and then type the following items in the large text box, one per line: NM , NM , D49487, NM Select an output format. The help text at the bottom of the Table Browser page describes the output formats. Not all options may be available for a given query. The All Fields...format displays the entire set of fields for each record in the output. The Selected Fields... format is useful when the user wishes to create output that contains only a subset of fields that will be used as input for further data processing or if the user desires to link in fields from an associated table (step 7). The Sequence option returns the sequence underlying the annotation in Using Biological Databases Current Protocols in Bioinformatics

20 Figure Output from the Table Browser query described in Support Protocol 2, steps 4 to 6, showing regions of chromosome 7 in the Feb (GrCh37/hg19) human genome assembly associated with the identifiers NM , NM , D49487, and NM The UCSC Genome Browser FASTA format. The GTF, BED, and custom track options are useful for saving the output into a format that can be displayed as a custom track in the Genome Browser. The Data Points format, which is available only for wiggle and Conservation tracks, is useful for displaying the conservation scores associated with individual base locations; in contrast, the Conservation track s MAF format displays the multiple species alignments underlying the conservation scores. To display a set of search results in the Galaxy genome analysis tool, check the Send Output to Galaxy box. For this example, choose the Selected Fields...output format. 6. Click the Get Output button to submit the query and display the results. By default, the Table Browser displays the query output in the user s Web browser. To save the data to a file on the local computer, type a file name in the Output File text box, and select the plain or compressed file type option before clicking the Get Output button. Many output formats including the Selected Fields... format used in the example require an additional setup step before the output is displayed. On the setup page associated with the example here, check the Name, Chrom, txstart, and txend boxes in the section labeled Select Fields from hg19.knowngene, and then click the Get Output button. The Table Browser will display the output shown in Figure Link in additional data from tables associated with the table being queried. The Linked Tables feature included on the Selected Fields... output format setup page provides a convenient way to pull in data from additional tables without having to conduct multiple queries. It is easy to expand the query outlined in the previous step to display additional data associated with the selected genes by linking in the associated tables. The kgxref table, linked by default when the UCSC Genes track is selected, provides a convenient crossreference among gene IDs and information from several different sources such as RefSeq, Swiss-Prot, HGNC, etc. For this example, return to the field selection setup page in step 6. At the top of the page, check the Name, Chrom, txstart, and txend boxes in the hg19.knowngene section as before. Move down to the hg19.kgxref section and check the GeneSymbol and Refseq fields to add this information to the output. Scroll down to the Linked Tables section; check the box on the hg19 kgalias and then click the Allow Selection from Checked Tables button at the bottom of the page to open the query to information in the kgalias table. The Table Browser will redisplay the page with the hg19.kgalias table added. Check the Alias field in the hg19.kgalias section, and then click the Get Output button. The Table Browser will display a comma-separated list of aliases, followed by the HGNC gene symbol and the RefSeq accession associated with each UCSC Genes record in the output shown in step Click the Summary/Statistics button to display a table of basic statistics about the current query. Current Protocols in Bioinformatics